228 THE CANADIAN ENTOMOLOGIST. 



tude resembles figure 6, the dorsum on last segments being convex, the 

 head turned in opposite S and 9, the lowest part of the curve being at 

 6th. Two or three hours later the body hangs straight, and the four 

 anterior segments are bent almost at a right angle to the others. The 

 head continues to droop, and by this it is made certain that the final 

 change approaches. Presently there is a twitching of the spines, first con- 

 fined to one segment, but extending soon over the whole body, and 

 changing into a waving motion. This is accompanied by a twisting of the 

 segments beneath the skin, which increases in strength and continues some 

 minutes. Two or three times a spasm of contraction comes on by which 

 the body is lifted up into the last one or two segments and let fall again. 

 Then a creeping movement under the skin commences, extending from 

 the posterior segments forward, and seems to break the skin loose from 

 the body, and one wave after another runs along till the distended skin on 

 the anterior segments bursts. This always takes place on the middle of 

 the dorsum, on the 3rd segment, and the mesonotum of the chrysalis is 

 forced through, splitting the skin up to the head (or first segment), and 

 sometimes splitting the skin of the head also. By the continued 

 creeping movement the body is slowly forced through the rent. As 

 this is oblique, the ventral side of the chrysalis is fully three seg- 

 ments behind the dorsal in the divesting, the skin on the anterior seg- 

 ments fitting tight as a glove, although it is loosening and packing in a 

 mass about the anal feet. In about 90 seconds from the time of the rup- 

 ture the skin on dorsal side has been pushed back to 10, and the effort 

 begins for the extrication of the tail of the chrysalis from the caterpillar 

 skin. This tail must be withdrawn and fastened outside the skin to the same 

 button of silk which the caterpillar clung to. At this instant the skin 

 covers the ventral side of the chrysalis to 8th segment, but is moving up 

 constantly, and as the chrysalis bends the posterior half of the abdomen 

 sharply back to force the tail out of the sheath, the segments 

 are pinched together and there is at the same time a pinching in of the 

 skin. But there is no seizing of the outside of the skin ; if there were 

 no other reason, the spines would make this impossible. The tail now 

 free, the chrysalis straightens itself up, and swinging on the ligament, lifts 

 itself towards the silk, the last segment describing an arc of a circle of which 

 the ligament is the radius, and the tail, which at the same instant is curved 

 forward, is brought round and over the considerable packet of the old skin 

 and with precision strikes the silk. An observer, knowing nothing of the 



